Research
We apply methods from computational learning, machine learning, data mining, database algorithmics and formal logic to the design of innovative algorithms and systems for data analysis and prediction.
Our mission is to develop practical solutions from sound principles.
Research at the group can be roughly divided into four lines:
1. Data mining and machine learning
- predictive and explanatory models
- recommender systems
- knowledge discovery in databases
- computational learning theory
- analysis of structured, non-relational data: sequences, trees, graphs, linguistic data, XML analysis
People involved: Marta Arias, Argimiro Arratia, José L. Balcázar, Santiago Boza, Jorge Castro, Ricard Gavaldà, Josefina López, Gilles Blondel, Matteo Ruffini, Marie Ely Piceno
2. Massive data analysis with emphasis on data streams
- efficient algorithms, scalability
- real-time analytics
- mining with limited resources
People involved: Marta Arias, José L. Balcázar, Jorge Castro, Ricard Gavaldà, Jordi Delgado
3. Applications of data mining techniques to specific domains
- social network analysis
- finances
- healthcare data
- utility data
- computer system performance analysis
- transportation networks
- sports analytics
- ecological data
People involved: Argimiro Arratia, Marta Arias, Jaume Baixeries, Jorge Castro, Ricard Gavaldà, Rafael Mena, Javier Fernández, Maria Slanova, Toni Rodríguez, Martí Zamora
4. Mathematical linguistics
- logic-based approaches to linguistic analysis
- quantitative linguistics
- language evolution and acquisition
People involved: Glyn Morrill, Ramon Ferrer, Jaume Baixeries, Bernardino Casas, Josefina Sierra, Toni Hernández, Oriol Valentín, Carles Cardó